CircleSearch：流媒体VoD环境下的P2P搜索算法

在基于P2P的流媒体VoD环境下，搜索机制是影响服务质量的关键因素之一。该文提出一种新的基于环状网络的分布式自适应概率搜索算法——CircleSearch。它依据节点本体内容生成结构化的环状拓扑网络，保证算法的可扩展性、健壮性和分布性。自发组织基于本体距离和连接可靠性的small world覆盖网络，减小消息的转发次数，采用自适应搜索算法提高文件定位的准确性。仿真实验表明，CircleSearch以其高成功率、低服务器负载和很小的平均搜索长度实现媒体流传输的即时、稳定、可控和连续，为VoD服务提供QoS保证。

CircleSearch: P2P Search Algorithm in Streaming Video-on-Demand Environment

In the environment of streaming Video-on-Demand(VoD) over P2P, search scheme is one of the key factors affecting the Quality of Service(QoS). This article proposes a new self-adaptive probability based search algorithm over circle networks——CircleSearch, which creates circle networks based on the contents of nodes to guarantee the scalability, robustness and distribution of the algorithm. This algorithm spontaneously constructs the “small world” overlay network based on the distances and link reliabilities, which significantly decreases the number of transmission. It uses self-adaptive probability based search algorithm to increase the accuracy of file location. Simulation results show that CircleSearch can provide real-time, stable, controllable and continuous media streaming for system and guarantee the QoS of VoD services with small average path lengths, high success rates and low servers’ loads.